AI is the buzzword marking the technological advancements in every industry, where Oil and Gas is no exception. It is true that the dynamic nature of operations, largely dealt with subsurface conditions, poses a great deal of difficulty for surface operators in producing field & contractors in drilling to implement these new innovations. But the evolving data science, going hand by hand with the domain, is causing a great shift of acceptance among the Oil and Gas personnel. Going by the words of EY, by 2021, the investments in AI by the Oil & Gas will be around $52.2 billion per year, which is currently just $19.1 billion.
Talking about the upstream production phase of this industry, field engineers are finding it difficult to manage the wells efficiently. The current scenario is such that if you considering a big operator having around thousands of producing wells spread across various fields of US, the operator is facing the issue of determining the health of each well. It is true that real-time remote monitoring feature has solved most of the challenges which posed earlier but for field engineers working from an operations centre managing around 100 wells each, it will be a challenge to determine the health of each well accurately.
In today’s world, the field engineers have trained eyes which could look upon the signatures of the well parameters & determine the health of each well. But how exactly the whole process of determining the health of a well & field visit to that well take place? The field engineer will commence the process by having a quick glance at the signatures of well & field parameters like production rate, casing pressure, tubing pressure, line pressure of the pipeline connected to different wells, etc. With their experience, they could easily detect the variations in the data & make up the mind as to whether this well needs attention or not. So, after spending around 3 hours to look through the parametric data of 100 wells under his jurisdiction, the field engineer will decide critical wells to visit & according to that will plan out the route for the day to visit marked wells & resolve the issues occurring at the wellhead.
Already spending 3 hours of the day just for planning, the field engineers looked back at the result where, they had missed the critical wells or took the good wells into account. There were wells which might not have needed visit & time could’ve been saved or even the wells could’ve marked wrongly resulting in non-visit to low health wells which have a negative impact to the operator both in terms of loss of time & thus money.
Applying different features just proved to be a slight increase in planning but Production Performance Digital Assistant proved to be a game changer for this leading operator of US. Producing gas from the around 2100 wells spread across the fields of US, the operator faced the same challenge discussed having the artificial lift installed in most. From the past two years, the operator saw that inefficiency of well health determination & planning caused about NPT of about 20 days per year. The loss of production & well intervention to resolve the issue of liquid loading summed up to the overall loss of millions for the operator.
So how this Digital Assistant made life easier for this operator? Digital Assistant uses machine learning data models trained on the historical signatures of the operations data which consists of production rate, casing pressure etc., & the heuristics rules used by the field engineer to determine the signatures. After ingesting the data, the app shows the score of each well to the field engineer, the least performing well is shown first. The score is also provided simultaneously for every well to have in-depth information about a well. An additional feature is also in the package with the digital assistant where the engineer can have a field tour plan in place covering most of the scored wells to reduce the time for planning the field visit.
Talking in numbers of the impact that the Production Performance Digital Assistant had on this operator, the process of determination of well health & planning was completely automated which led to time-saving of about 2.5 hours a day. Also, the confidence of the model was recorded at about 99.75% which assured that the model was nearly right all the time which led to reduced loss of production & intervention applied to the well. With the plan already in place to pick as to which route to take to visit the wells needing attention, the operator recorded a whopping savings of about $10 million last year!
Do you need any assistance in determining your well’s health or even planning the field trip? Contact us today so that you can get a taste of how this Digital Assistant will aid you in saving time with reduced loss of production.
Let us know if you liked this post & stay tuned for our next post to know how we averted 4 large oil spills for an operator using our feature of Tank Intelligence.
SUBSCRIBE TO Flutura
Vivamus suscipit tortor eget felis porttitor volutpat. Praesent sapien massa, egestas non nisi.